revision 02

tex_review/chapters/experiments.tex

@@ -101,13 +101,14 @@ Nevertheless, Android app and offline application are both use the same C++ back
 %Sensor measurements are recorded using a simple mobile application that implements the standard Android SensorManager. 
 
 The experiments are separated into five sections:
-At first, we discuss the performance of the novel transition model and compare it to a grid-based approach.  
+At first, we discuss the performance of the novel transition model and compare it to our previous approach using a gridded graph structure.  
 In section \ref{sec:exp:opti} we have a look at \docWIFI{} optimization and how the real \docAPshort{} positions differ from it. 
 Following, we conducted several test walks throughout the building to examine the estimation accuracy (in meter) of the localization system and discuss the here presented solutions for sample impoverishment.
 \add{In section \ref{sec:eval:act} the threshold-based activity recognition is evaluated, providing a detection rate for the test walks utilized before.}
 Finally, the respective estimation methods are discussed in section \ref{sec:eval:est}.
 
 \subsection{Transition}
+\label{sec:exp:transition}
 
 \begin{figure}[t]
 	\centering
@@ -171,9 +172,9 @@ For example walking through a door, would result in a strong reduction of differ
 If the state evaluation is then used to assign weights to particles, the crucial problem of sample degeneracy often occurs. 
 With a mesh, on the other hand, walkable destinations can also be located in a room behind a wall.
 In combination with the continues movement, this allows for a high versatility of particles even in such situations. 
-Another method to fix the problems shown in fig. \ref{fig:transitionEval:d}, is by adding an activity recognition (walking up, down straight) or to incorporate a barometer. 
-Nevertheless, in most cases it is an advantage if a sensor model delivers good results on its own, without further dependencies. 
-For example, if a sensor is currently unavailable or damaged, the system is still able to provide prober results. 
+Another method to fix the problems shown in fig. \ref{fig:transitionEval:d}, is by adding an activity recognition (walking up, down, straight) or to incorporate a barometer. 
+Nevertheless, in most cases it is an advantage, if a sensor model delivers good results on its own, without further dependencies. 
+For example, if a sensor is currently unavailable or damaged, the system is still able to provide proper results. 
 
 Besides the advantages the mesh offers, it also has a few disadvantages compared to the graph.
 The computation time has increased due to the calculation of reachable destinations. 
@@ -299,6 +300,13 @@ In contrast, the $D_\text{KL}$-based method extends the transition and thus uses
 We set $l_\text{max} =$ \SI{-75}{dBm} and $l_\text{min} =$ \SI{-90}{dBm}.
 For a better overview, we only used the KDE-based estimation, as the errors compared to the weighted-average estimation differ by only a few centimeter. 
 
+\addy{The same applies for an accuracy comparison between the graph-based model and the navigation mesh as part of the overall system.
+Both provide very similar localization errors regarding the conducted walks.
+This is not a big surprise, as the accuracy of the pedestrian’s position based on the estimated state and thus the complete posterior density (weighted particle set).
+It is obvious, that choosing a graph with a grid-size of e.g. \SI{2}{} x \SI{2}{\meter} would worsen the results.
+This leads to the statement, that the approximation error of walking alongside the edges of a (reasonable sized) gridded graph is small enough that it has no significant influence on the overall localization accuracy compared to a true continuous motion. 
+Nevertheless, as shown in section \ref{sec:exp:transition}, the navigation mesh offers several major benefits by highly reducing the memory footprint.}
+
 \begin{table}[t]
 	\centering
 	\begin{tabular}{rrrrcrrrcrrr}